IS480 Team wiki: 2017T1 LeGIT FinalWiki
Contents
- 1 Project Progress Summary
- 2 Project Management
- 3 Quality of Product
- 4 Reflection
Project Progress Summary
Final Slides: Final Slides
Video Pitch: Link to Google Drive
Project Highlights
Highlights |
---|
Implementation of Google VR View, enabling VR Technology on the ecommerce website! |
Developed additional VR Application with Unity for Android platforms, featuring VR Interactions! |
Enhanced Ecommerce setup and plugin with Google Analytics. User shopping behavior analysis and checkout process drop-off rates. |
Machine Learning with Apache Mahout. Product recommendation for both logged in users and guests with K-Nearest Neighbor (kNN) algorithm and Pearson Correlation Similarity. |
Project Management
Project Status
Project Schedule (Planned vs Actual)
Due to some unforeseen events happening during the project phase, the schedule has been altered slightly. Change requests such as the inclusion of Multiple Address Lines for customers to select from and requirements for a promotional code system added new functionalities. The incident of data misalignment with business requirements also delayed development by a week, sending some functionalities due for completion in Sprint 6 into the product backlog. However time was also saved through the implementation of Google Analytics Enhanced E-commerce Plugin which took on the burden of session and event tracking analytics.
After the midterms, additional events happened. Such as the change request to include Product Recommendation to the website was accepted by LeGIT in view of the benefits it brings to the business and experience it gives us by using Machine Learning. Unexpected technical difficulties such as Angular platforms difficulty retrieving data from Google Analytics also made us turn to customizing Google Analytic's website itself for Sponsor's use.
Lastly, the team also discovered the additional features of creating a VR application (Android) with Unity to allow interaction with the VR itself to enhance the VR experience. These events and changes all lead to controlled changes to our schedule, in order to deliver the most satisfactory product to our Sponsor.
Highlighted in red, are the changes to our schedule
DePICT Project Management
Methodology
MoSCoW Scope Prioritization
Through the use of the MoSCoW Prioritization method, Team LeGIT keeps close track over the scope requirements of the project as well as catering for possible inclusion of additional function requests that arises throughout the project.
Sprint Velocity Chart
Based on the planned story points given to all our product backlogs. All 11 Sprints completing an average of 90 Story Points will ensure successful project delivery. For the purpose of planning ahead, end of sprints go through Sprint Retrospective to plan with achieving the maximum number of Story Points possible the next Sprint.
Risk Management
Issue Management
Change Management
Technical Complexity
Google Analytics (Enhanced Ecommerce)
The simple and general way of employing Google Analytics (page tracking)
The Team LeGIT advanced Enhanced Ecommerce Plugin for Google Analytics enables tracking of product, checkout process, session movement information to Google Analytics to build Customer and Checkout Behavior analytics:
Results:
Machine Learning for Product Recommendation (Apache Mahout)
Apart from the technical complexity of understanding, collecting and deploying of Machine Learning Algorithm for Product Recommendation with Apache Mahout. Team LeGIT went one-step further to ensure direct connection to SQL database for the Apache Mahout instead of CSV file creation (lengthy and excessive computing power requirement). As well as including guest recommendations with PlusAnonymousConcurrentUserDataModel implementation in Mahout.
The normal Apache Mahout:
Team LeGIT's implementation of Apache Mahout with MYSQLDatabase connection and PlusAnonymousConcurrentUserDataModel for Guest Profile Recommendations
Virtual Reality (Embedding on Website + Android Application)
Probably Team LeGIT's most daring endeavor, we set off to implement Virtual Reality technology to probably the first retail outlet in Singapore to feature VR with its products in Brick-n-Mortar.
We implemented VR in not just one, but two areas. Embedded VR that appears on every Product Detail page while the customer is browsing or redirected to the Ecommerce website through QR Code in-store. Secondly, a VR application that can be downloaded onto an Android platform, that provides user the ability to interact and change the bedding patterns and colors within the VR viewing session itself.
These were all possible through extensive research and self-experience with software and technologies such as Google Sketchup, V-Ray, Google VR View and Unity. With this we have enabled users without VR devices to view a 360 Panorama on the website, users who has low-cost VR devices like Google Cardboard to view it in VR. And for users who own VR device like Samsung Gear VR, the ability to use the VR app to interact with the VR itself to choose and select product patterns and colors.
Quality of Product
Project Deliverables
Stage | Specification | Modules |
Project Management | Minutes | Minutes |
Metrics | Metrics | |
DePICT Project Management | DePICT | |
Risk Management | Risk | |
Issue Management | Issue | |
Change Management | Change | |
Requirements | Payment Gateway Research | PaymentResearch |
Analysis & Design | Business Process and Architecture Diagram | Diagrams |
Testing | User test plan | User Test 1 |
Handover Documents | Developer_and_AWS_Deployment_User_Guide | Developer and Server Guide |
User Manual for Admin Users | User Manual | |
VR Interactive Manual (Unity) | VR Interactive Manual | |
VR Rendering Manual (Sketchup + V-Ray) | LeGIT Youtube Video Guide |
Deployment
To View Application, visit:
E-commerce Portal: https://www.highlanderdeluxe.com/
Administrative Portal: [Officially Deployed for Sponsor use, please email yuxuan.tee.2014@sis.smu.edu.sg to request for access]
Testing
Functional Testing Done after every Iteration User Testing done
User Testing | Date | Venue | No. Of Users | Link |
---|---|---|---|---|
1 | 20th August | Home of User & SOL GSR B1-06 | 6 B2C Users, 2 Highlander London Staff | User Test 1 Link |
2 | 9th - 10th October | At Tester's Convenience | 15 B2C Users, 2 Highlander London Staff | User Testing 2 Link |
3 | 6th - 11th November | At Tester's Convenience | 31 B2C Users, 5 Highlander London Staff | User Testing 3 Link |
4 | 17th November | SOE Classroom 2-1, Tester's Convenience, Sponsor's Residence | 66 B2C Users (Including 48 SMU Students), 5 Highlander London Staff | User Testing 4 Link |
Reflection
Team Reflection